The margin of error is too high to say what the true value is (or should be), but in no way does that imply that the true value should be ZERO. A small number does not automatically mean zero.
You could make up similar lies and say that "statistically, the amount of lead in your drinking water is zero", and have it be well less than 12 parts in 10,000, and it's still non-zero, and at neurologically toxic levels. Just because a number is small does not mean that we cannot accurately measure it, nor does it mean that the expected value is zero. Claiming that the expected value is zero when there are twelve in your (large) sample is a horrible misuse of statistics.
The number of black women is 12, not 0. I don't understand what you wish to gain by egregiously twisting the numbers. Yes, it's a low percentage. No, it's not zero.
What do you wish to gain by being painfully pedantic? Black women are 5-6% of the population, but 0.01% of funded founders. That is a rounding error, effectively zero. Why not focus on the actual point of the post instead?
Because the title primed me to expect one thing, and then the post body rapidly contradicted it by providing twelve vetted examples of the class that the title claimed should not exist. My mind does not get over being misled and confused like that easily, especially when the fix to improve it is so trivial (like saying "virtually none" instead of "statistically zero").
The deal is that targeted pedantry like this completely derail the discussion, which is not 12 ≠ 0, but how and why black women are being excluded from tech.
I've got a great way of increasing the percentage by an order of magnitude! Doing the math correctly would immediately boost the number to 0.1% from your stated 0.01%. See, being painfully pedantic has already paid off handsomely.
(And don't worry, at least you were only off by a factor of 10. The original article is off by a factor of 100).
You could make up similar lies and say that "statistically, the amount of lead in your drinking water is zero", and have it be well less than 12 parts in 10,000, and it's still non-zero, and at neurologically toxic levels. Just because a number is small does not mean that we cannot accurately measure it, nor does it mean that the expected value is zero. Claiming that the expected value is zero when there are twelve in your (large) sample is a horrible misuse of statistics.